Cream-vented Bulbul Pycnonotus simplex of Borneo was previously considered to be polymorphic in iris colour, having either red or white (creamy-yellow) irides. Mitochondrial DNA sequence comparisons, however, indicate that white- and red-eyed Bornean individuals are not closely related to one another. Instead, white-eyed birds are sister to Ashy-fronted Bulbul P. cinereifrons of Palawan Island, in the south-west Philippines, and red-eyed birds are sister to white-eyed P. simplex of the Thai-Malay Peninsula. Consequently, we elect to treat the white-eyed Bornean population as a distinct, previously overlooked species. In respect to plumage, white- and red-eyed individuals are almost identical, varying only slightly in the amount of yellow coloration in their feathers. The two taxa are sympatric at some localities, but white-eyed individuals are rarer and more consistently associated with mature forest than red-eyed birds.
Cream-vented Bulbul Pycnonotus simplex is a uniformly olive-brown species that occurs from southern Indochina throughout the Sunda Islands, except Palawan in the south-western Philippines (Fig. 1). Its classification has been complicated by its variable iris colour (Hoogerwerf 1966, Mees 1986, Dickinson & Dekker 2002), which has led to a confused taxonomic history and disagreement concerning the validity of iris colour as a subspecific character (Hoogerwerf 1966, Mees 1986). Although as many as six subspecies of P. simplex have been named (Table 1), only three or four are currently recognised (Dickinson & Christidis 2014, Eaton et al. 2016, Clements et al. 2017). Here, we follow the classification of Dickinson & Christidis (2014) because of its thorough documentation. It lists three subspecies.
Nominate P. s. simplex Lesson, 1839, occurs from south-eastern Myanmar to Sumatra, the southern Natuna Islands and Borneo. Individuals on the Asian mainland and Sumatra have white to grey or yellowish-white irides (Meyer de Schauensee 1958, Mees 1986). Populations on Borneo, Billiton, and Bangka Islands have primarily red irides (Mees 1986), a distinction that caused Chasen & Kloss (1929) to treat them subspecifically as P. s. perplexus. Yet another subspecies, P. s. oblitus, was originally described from the southern Natuna Islands based on its red irides and supposedly longer, heavier bill (Deignan 1954). However, in a thorough review of the species' taxonomy, Mees (1986) noted that individuals with ‘white’ irides (i.e., pale yellow or creamy yellow) also occur occasionally on Borneo, often in close proximity to red-eyed individuals, and he also found no size difference between P. s. oblitus and specimens from Borneo. Mees agreed with Hoogerwerf (1966) that iris colour is not a reliable subspecific character in these bulbuls and treated P. s. perplexus and P. s. oblitus as junior synonyms of P. s. simplex. As a result, P. s. simplex now includes both red- and white-eyed individuals.
The second subspecies, P. s. prillwitzi E. J. O. Hartert, 1902, occurs on Java, but its iris colour was not noted in the original description. Later, Hoogerwerf (1996) indicated that Javan birds have red, orange or reddish-brown irides. This subspecies was described as brighter overall than other subspecies, with greener upperparts and more yellowish underparts (Hartert 1902). The third subspecies, P. s. halizonus Oberholser, 1917, which includes P. s. axanthinus Oberholser, 1932, as a junior synonym, occurs on the northern Natuna and Anamba Islands. It was described as having white irides and being larger than individuals from Borneo or the Thai-Malay Peninsula (Oberholser 1917).
Figure 1.
Map of Sundaland showing range of Cream-vented Bulbul Pycnonotus simplex with labelled type localities of subspecies from Table 1 and the collecting sites of P. simplex specimens whose ND2 was sequenced.
Recent work has shown that iris colour actually can be an important trait for differentiating bulbul taxa that previously were considered conspecific based on similarities in plumage and vocalisations (Garg et al. 2016). On Borneo, two common bulbul species that are similar in plumage and voice to each other and to P. simplex are readily distinguished by their eye coloration. Whereas P. simplex has dark red, creamy or white irides, Red-eyed Bulbul P. brunneus has a distinctly orange or two-toned iris (red at the edge and orange towards the centre) and Spectacled Bulbul P. erythropthalmos has a dark red iris surrounded by an eye-ring of yellow skin. Based on these observations, eye colour would appear to play a key role in recognition among some bulbul species.
In the course of genetic comparisons for ongoing studies of Bornean bird biogeography and phylogeography (e.g., Sheldon et al. 2015, Lim et al. 2017, Shakya et al. 2018), we discovered that red- and white-eyed P. s. simplex on Borneo are not members of the same species. They are not even sister taxa. White-eyed individuals—although remarkably similar to red-eyed birds in plumage—are members of a distinct species.
TABLE 1
Past and current names of Pycnonotus simplex populations.
Methods
We compiled a dataset of mitochondrial ND2 (NADH dehydrogenase subunit 2) sequences of 15 P. simplex individuals from Borneo (11 red-eyed and four white-eyed) and one from the Thai-Malay Peninsula (white-eyed), as well as outgroup sequences of four P. brunneus, two Ashy-fronted Bulbuls P. cinereifrons, nine Olive-winged Bulbuls P. plumosus, one P. erythropthalmos, one Straw-headed Bulbul P. zeylanicus and two Buff-vented Bulbuls Iole charlottae (Table 2). These outgroups were selected based on a recent phylogenetic study of the Pycnonotidae (Shakya & Sheldon 2017). Sequences were either newly generated or obtained from GenBank. For newly generated sequences, we extracted total genomic DNA from frozen or alcohol-preserved samples using a DNEasy Blood and Tissue Kit (Qiagen, Hilden, Germany) following the manufacturer's protocol. PCR amplifications of ND2 were performed using primer pairs L5215 (Hackett 1996) and HTrpC (STRI) as described in Shakya & Sheldon (2017). Sequences were aligned using MUSCLE (Edgar 2004), and maximum-likelihood (ML) tree searches were conducted using RAxML v8 (Stamatakis 2014) with the GTRGAMMA model for three partitions corresponding to each codon position. A total of 1,000 non-parametric bootstraps was run to obtain branch support values in the best tree generated from RAxML.
We also examined and measured skin specimens in the collections of the Louisiana State University Museum of Natural Science (LSUMNS), Baton Rouge, LA, and the Western Foundation of Vertebrate Zoology (WFVZ), Camarillo, CA, USA. Specifically, we measured bill length (from the nares to the tip), bill width (at nares), bill depth (at nares), wing-chord length, tail length and tarsus length. We did not re-examine type specimens of P. simplex subspecies because original or revisional descriptions of subspecies outside Borneo distinguished them from Bornean red- and white-eyed individuals either by plumage colour or size (Oberholser 1917, Hoogerwerf 1966, Mees 1986). Also, plumage colour in specimens of P. simplex changes rapidly with age (Hoogerwerf 1966) and, thus, re-examination of plumage at this stage would be uninformative.
TABLE 2
Specimen, tissue, and GenBank numbers of taxa included in phylogenetic analyses.
Results
Genetic comparisons indicate that white- and red-eyed P. simplex from Borneo do not form a monophyletic group (Fig. 2). Red-eyed Bornean birds are sister to white-eyed birds from the Thai-Malay Peninsula, whereas white-eyed Bornean birds are sister to P. cinereifrons of Palawan. ND2 sequences of white-eyed Bornean birds differ by a mean 14% from red-eyed Bornean birds and 12% from P. cinereifrons. Compared to these relatively large distances, ND2 of red-eyed Bornean birds differs by just 4% from white-eyed P. simplex of the Thai-Malay Peninsula.
Mensural data from museum skins indicate similarity in most dimensions between red-eyed and white-eyed individuals (Table 3). Wing, tail and tarsus lengths did not differ significantly. However, red-eyed birds appear to be slightly heavier than white-eyed birds: red, mean 20.7 ± 2.0 g (n = 11); white, mean 17.8 ± 2.4 g (n = 12). Red-eyed birds also appear to have slightly larger bills in all dimensions. No incidental biological circumstances appear to cause the size differences; e.g., the mass variation does not result from more breeding females in one taxon than the other.
Figure 2.
Maximum likelihood tree generated from mitochondrial ND2 sequences. Numbers adjacent to branches indicate bootstrap support. Illustrations of heads of each species are shown to the right (depictions by Subir B. Shakya).
TABLE 3
Measurements of the new species and P. s. simplex from Borneo.
Discussion
Taxonomy.—Genetic comparisons using mitochondrial ND2 sequences reveal that white- and red-eyed individuals of P. simplex from Borneo are not members of the same species. Their segregation by iris colour is backed by a deep ND2 sequence divergence (14%). Red-eyed Bornean birds are sister to white-eyed P. s. simplex of the Thai-Malay Peninsula at 4% ND2 divergence, whereas white-eyed Bornean birds are sister to P. cinereifrons of Palawan at 12% ND2 divergence. The perfect association of ND2 haplotype with iris colour across the region where the two morphotypes co-occur suggests assortative mating and the absence of mitochondrial gene introgression. For the red-eyed Bornean population, iris colour difference is adequate for recognition as a separate subspecies (P. s. perplexus; Chasen & Kloss, 1929, Eaton et al. 2016), or possibly even species, from white-eyed P. s. simplex of the Thai-Malay Peninsula and Sumatra. Because the two taxa are relatively closely related in respect to ND2 distance, species-level reclassification should depend on future research into gene flow and song. Plumage variation is unlikely to be helpful because of the similarity between these populations and the tendency for colours to change in museum specimens (Hoogerwerf 1966, Mees 1986). For the white-eyed Bornean population, its large genetic distance from P. cinereifrons and distinct morphology (Fig. 2) are sufficient to recognise it as a species.
There are no other taxa whose names apply to the Bornean white-eyed population. P. simplex subspecies have been described from Sumatra, the Anamba Islands, the northern Natuna Islands, the southern Natuna Islands, Java, and northern Borneo (Table 1, Fig. 1). Iris colour and size connect white-eyed populations of Sumatra and mainland Asia in subspecies simplex (Fig. 1). Larger size distinguishes the white-eyed subspecies halizonus (including its junior synonym axanthinus) of the Anamba and northern Natuna Islands from other white-eyed populations: e.g., mean wing length for halizonus 83.6 mm (n = 4) vs. 74 mm (n = 12) for Bornean white-eyed individuals. The red-eyed subspecies on the southern Natuna Islands (oblitus) belongs with Borneo's red-eyed population based on iris colour, size and plumage (Mees 1986). Eye colour of the Javan population (prillwitzi) varies from red to orange or red-brown, and its plumage is more distinctive than any other population of P. simplex, including red- and white-eyed Bornean populations (Hoogerwerf 1966). Apart from these taxa of P. simplex, there is no evidence that early taxonomists confused other species of similar-looking bulbuls with the white-eyed Bornean population of P. simplex. Pycnonotus brunneus has two subspecies: P. b. brunneus (including P. b. zaphaeus) and P. b. zapolius. Both of these were described as having red irides. P. erythropthalmos (including P. e. salvadorii, sometimes recognised for Borneo) has a red iris with a yellow eye-ring. Because no name applies to the Bornean white-eyed population, we name it:
Pycnonotus pseudosimplex, sp. nov.
Cream-eyed Bulbul
Holotype.—Study skin, Louisiana State University Museum of Zoology (LSUMZ 188187), Baton Rouge; tissue, LSU Museum of Natural Science (LSUMNS B-84971); adult male; Malaysia, Sarawak, Miri Division, Miri, Lambir Hills National Park, 04°12′N, 114°25′E; elevation c.120 m (from Google Earth); habitat: secondary forest continuous with old-growth hill forest; collected 23 April 2013; prepared by Frederick H. Sheldon (FHS), original catalogue number FHS 1022; photo by John Mittermeier (Fig. 3A); GenBank accession no. MK298058.
Diagnosis.—The iris of P. pseudosimplex is creamy yellow instead of crimson-red as in Cream-vented Bulbul P. s. simplex of Borneo. Its plumage also differs from that of P. s. simplex in being yellower on the throat and vent, creating greater contrast with the darker breast and flanks. From P. cinereifrons, P. pseudosimplex differs in being much smaller: P. cinereifrons mean 31.7 g (n = 3); P. pseudosimplex mean 20.7 g (n = 11). P. cinereifrons also has an olive tinge to the leading edge of its remiges, which is especially obvious on the folded wings of specimens (similar to P. plumosus of Borneo). From P. brunneus, P. pseudosimplex differs in having a white iris instead of an orange or two-toned iris. Similarly, P. erythropthalmos differs in having a dark red iris and a circle of yellow skin around the eyes.
Description of holotype.—Approximate colour names used for description, with standard colour names from Ridgway (1912) and alphanumeric colour codes from Munsell (1950) in parentheses. Crown, nape and back uniform olive-brown (Olive Brown; 5Y3/4) becoming yellower on rump and uppertail-coverts (Dresden Brown; 2.5Y4/4). Tail dark brown (Chestnut Brown; 10YR3/2). Cheeks similar in colour to crown (Olive Brown; 5Y3/4) with paler shaft-streaks to feathers. Throat pale yellow (Olive Buff; 5Y8.5/4); breast slightly darker than throat with darker olive-brown diffuse streaks (5Y4/2). Flank feathers olive (Light Brownish Olive; 2.5Y5/6). Undertail feathers cream-coloured (Naples Yellow; 5Y7/6). Wing like tail (Chestnut Brown; 10YR3/2) with olive-yellow edges. Soft part colours recorded on collection: bill black; iris creamy yellow; legs and toes dark brown. No moult; testis 2 × 1 mm, skull: 100% ossified; stomach: full of fruit with 1 mm-long seeds.
Measurements of type.—See Table 3.
Paratypes.—We designate 11 paratypes, which include specimen LSUMZ 181671 and ten specimens held at WFVZ: 38742, 38743, 38746, 38747, 38754, 38775, 38776, 38778, 38780, 38782 (listed in Table 3). The paratypes were collected in various parts of Sabah, Malaysian Borneo (unlike the holotype, which was taken in Sarawak, Malaysian Borneo). Measurements of all paratypes and their iris colours are reported in Table 3.
Etymology.—For more than 100 years this species has been confused with P. simplex, i.e., since Carl Lumholtz collected the first white-eyed specimen in ‘Boeloengan’ (Balungan Regency) along the Kayan River in eastern Kalimantan, Indonesian Borneo, in 1914 (Lumholtz 1920, Voous 1961). Apart from eye colour, the two species are almost indistinguishable; hence, we refer to the new species as ‘false’ or ‘pseudo’ simplex. The common name, Cream-eyed Bulbul, describes the main identifying character of the new species. It also evokes a time when P. simplex was known as the White-eyed Brown Bulbul, as opposed to the Red-eyed Brown Bulbul P. brunneus of Borneo (Smythies 1960).
Habitat.—P. simplex and P. pseudosimplex appear to be sympatric in mature forest interior and edge. Individuals of both species whose DNA was compared in this study were collected in the same localities during recent years: viz., Ulu Kimanis, Crocker Range National Park, Sabah, in 2008; Lambir Hills National Park, Sarawak, in 2013; and Batang Ai National Park, Sarawak, in 2018 (Fig. 1, Table 2). Both taxa were also mist-netted together in the early 1980s elsewhere in Sabah: at Bole River, Saliwangan Baru, and Mangkatai (Megatai) (localities described in Sheldon 2015). Although previous authors have stated that white- and red-eyed individuals co-occur throughout Borneo (Hoogerwerf 1966, Mees 1986), our experience and specimen records suggest that white-eyed individuals are considerably rarer and perhaps more habitat-restricted. P. simplex occurs to 1,100 m (Kelabit Highlands) and in several forest types of varying quality: good-soil dipterocarp, kerangas and peatswamp; primary, secondary, and heavily disturbed forest; and plantations. We have found P. pseudosimplex (only by mist-netting) to 500 m (Ulu Kimanis) near the edge of mature good-soil dipterocarp, and in kerangas forest nearer to sea level.
Morphology and voice.—Given their distinct ancestry, the striking similarity in plumage between P. simplex and P. pseudosimplex is remarkable. This is particularly true in specimens, less so in photographs (Fig. 3). They also closely resemble other species of brown bulbuls on Borneo to which they are not especially closely related: Red-eyed Bulbul and Spectacled Bulbul (Fig. 2). Perhaps these species are strongly selected for the cryptic advantage of dull brown plumage, but that observation does not hold for other bulbuls that occasionally occupy forest-edge habitat, such as Black-headed P. atriceps and Yellow-vented Bulbuls P. goiavier, which are both much more brightly coloured.
We are unaware of any vocal differences between P. simplex and P. pseudosimplex, although we would expect some. P. simplex is far more common than P. pseudosimplex and thus more likely to be represented in sound archives, but without accompanying iris colour data it is currently impossible to know which species is vocalising in recordings. In general, all of Borneo's brown bulbuls have similar songs that are not as distinctive (to the human ear) as Black-headed, Yellow-vented, or other species of bulbuls. Although vocalisations are certainly important, eye colour might be the main trait for species recognition among brown bulbuls. This has been observed in Streak-eared Bulbul P. blanfordi of Indochina and the Thai-Malay Peninsula (Garg et al. 2016). A large mitochondrial distance occurs between two vocally and visually similar subspecies, P. b. blanfordi and P. b. conradi, which are also most easily distinguished by iris colour: P. b. blanfordi has dark red irides, whereas P. b. conradi usually has dark grey irides (Garg et al. 2016). On Borneo, field comparisons of P. pseudosimplex with other brown bulbuls—especially how the species use their markedly distinct eye colours—are likely to yield rich information on the behavioural ecology of an otherwise generally under-appreciated group of birds.
Acknowledgements
We thank the Malaysian Prime Minister's Department and government agencies in Sabah and Sarawak for research permission, including specimen collection, and assistance. We also thank colleagues and students at Sabah Museum, Sabah Park, Sarawak Forestry Corporation, University Malaysia Sarawak, and Louisiana State University for help collecting specimens. Nomenclature in this paper was reviewed by the Working Group on Avian Nomenclature of the International Ornithologists' Union. We appreciate the help of two referees, Frank Rheindt and Frank Steinheimer, whose suggestions improved the manuscript and its adherence to the ICZN Code. The project protocol was approved by LSU's Institutional Animal Care and Use Committee. Funding was provided by the Coypu Foundation, National Geographic Society (8753-10), National Science Foundation (DEB-0228688 and DEB-1241059 to FHS, DEB-0743491, DEB-1557053 and DEB-1241181 to RGM), Global Genome Initiative Grant (GGI-LimH-2016) to HCL, the National Museum of Natural History Division of Bird Wetmore Fund, and a Smithsonian Institution Molecular Evolution Fellowship.
